--------------------------------------------------------------------------------------------------------------------------------
      name:  <unnamed>
       log:  C:\Users\Dell\Dropbox\OutcomeUncertainty\II Replication Folder\Nuclear Proliferation Log.log
  log type:  text
 opened on:  23 Sep 2017, 18:56:49

. clear

. use "Nuclear Proliferation Data.dta"

. *********************************************
. ******SUMMARY STATISTICS********************* - OK
. *********************************************
. su  pursueonly gjprog W_SystemRELa reg5_w persdumjlw_lag  land 

    Variable |        Obs        Mean    Std. Dev.       Min        Max
-------------+---------------------------------------------------------
  pursueonly |      7,109    .0323534    .1769492          0          1
      gjprog |      7,102    .0374542    .1898855          0          1
W_SystemRELa |      7,421    .2116745    .0108537   .1887663   .2403361
      reg5_w |      7,384    .3877031    .0878193   .2799952   .6291364
persdumjlw~g |      5,659    .2009189    .4007229          0          1
-------------+---------------------------------------------------------
        land |      7,422    3.130962    2.404144          0         20

. 
. 
. 
. *************************************************************
. ********************TABLE 2********************************** -OK
. *************************************************************
. eststo clear 

. xtlogit pursueonly persdumjlw_lag  land   timeSW timeSW2 timeSW3, nolog

Random-effects logistic regression              Number of obs     =      5,338
Group variable: ccode                           Number of groups  =        173

Random effects u_i ~ Gaussian                   Obs per group:
                                                              min =          1
                                                              avg =       30.9
                                                              max =         55

Integration method: mvaghermite                 Integration pts.  =         12

                                                Wald chi2(5)      =     143.67
Log likelihood  =  -210.8614                    Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
    pursueonly |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
persdumjlw_lag |   2.960682   .6351128     4.66   0.000     1.715883     4.20548
          land |   .8593746   .2006844     4.28   0.000     .4660403    1.252709
        timeSW |  -1.161097   .1169995    -9.92   0.000    -1.390412   -.9317825
       timeSW2 |   .0525975   .0067674     7.77   0.000     .0393336    .0658615
       timeSW3 |  -.0006246   .0001021    -6.12   0.000    -.0008247   -.0004245
         _cons |  -10.34679   1.496809    -6.91   0.000    -13.28048   -7.413101
---------------+----------------------------------------------------------------
      /lnsig2u |   3.103335   .2716499                      2.570911    3.635759
---------------+----------------------------------------------------------------
       sigma_u |   4.719334   .6410032                      3.616316    6.158786
           rho |   .8712984   .0304621                      .7990013    .9201886
--------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 166.84                 Prob >= chibar2 = 0.000

. qui estimate store nullSW

. xtlogit pursueonly W_SystemRELa persdumjlw_lag  land   timeSW timeSW2 timeSW3, nolog

Random-effects logistic regression              Number of obs     =      5,338
Group variable: ccode                           Number of groups  =        173

Random effects u_i ~ Gaussian                   Obs per group:
                                                              min =          1
                                                              avg =       30.9
                                                              max =         55

Integration method: mvaghermite                 Integration pts.  =         12

                                                Wald chi2(6)      =     145.88
Log likelihood  = -204.00725                    Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
    pursueonly |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
  W_SystemRELa |  -50.17194   14.62579    -3.43   0.001    -78.83796   -21.50592
persdumjlw_lag |    2.76741   .6244856     4.43   0.000     1.543441    3.991379
          land |   .9482929   .1923574     4.93   0.000     .5712794    1.325306
        timeSW |  -1.221398    .120665   -10.12   0.000    -1.457897   -.9848987
       timeSW2 |   .0545218   .0068467     7.96   0.000     .0411025    .0679411
       timeSW3 |  -.0006356   .0001037    -6.13   0.000    -.0008388   -.0004325
         _cons |  -.7154484   3.320279    -0.22   0.829    -7.223075    5.792178
---------------+----------------------------------------------------------------
      /lnsig2u |   3.313633   .2245033                      2.873615    3.753652
---------------+----------------------------------------------------------------
       sigma_u |   5.242595   .5884899                      4.207243    6.532736
           rho |   .8930982   .0214342                      .8432706     .928429
--------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 140.57                 Prob >= chibar2 = 0.000

. qui estimate store W_SystemRELaSW

. xtlogit pursueonly reg5_w  persdumjlw_lag  land   timeSW timeSW2 timeSW3, nolog

Random-effects logistic regression              Number of obs     =      5,324
Group variable: ccode                           Number of groups  =        173

Random effects u_i ~ Gaussian                   Obs per group:
                                                              min =          1
                                                              avg =       30.8
                                                              max =         55

Integration method: mvaghermite                 Integration pts.  =         12

                                                Wald chi2(6)      =     139.48
Log likelihood  = -208.57261                    Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
    pursueonly |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
        reg5_w |  -10.01523   5.591433    -1.79   0.073    -20.97424    .9437758
persdumjlw_lag |   3.105664   .6830799     4.55   0.000     1.766852    4.444476
          land |    .906495   .2332235     3.89   0.000     .4493853    1.363605
        timeSW |  -1.144654   .1181002    -9.69   0.000    -1.376126   -.9131815
       timeSW2 |   .0515397   .0067596     7.62   0.000      .038291    .0647883
       timeSW3 |  -.0006048   .0001002    -6.04   0.000    -.0008012   -.0004085
         _cons |   -7.63486   2.227774    -3.43   0.001    -12.00122   -3.268502
---------------+----------------------------------------------------------------
      /lnsig2u |   3.410856   .3000548                       2.82276    3.998953
---------------+----------------------------------------------------------------
       sigma_u |   5.503742   .8257122                      4.101611    7.385189
           rho |    .902032   .0265159                      .8364314    .9431123
--------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 170.31                 Prob >= chibar2 = 0.000

. qui estimate store reg5_wSW

. xtlogit gjprog  persdumjlw_lag land timeGJ timeGJ2 timeGJ3, nolog

Random-effects logistic regression              Number of obs     =      5,337
Group variable: ccode                           Number of groups  =        173

Random effects u_i ~ Gaussian                   Obs per group:
                                                              min =          1
                                                              avg =       30.8
                                                              max =         55

Integration method: mvaghermite                 Integration pts.  =         12

                                                Wald chi2(5)      =     162.77
Log likelihood  = -198.17231                    Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
        gjprog |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
persdumjlw_lag |   3.297021   .6541079     5.04   0.000     2.014993    4.579049
          land |    1.05633    .232921     4.54   0.000      .599813    1.512847
        timeGJ |  -1.558238   .1691096    -9.21   0.000    -1.889687   -1.226789
       timeGJ2 |   .0888615   .0127182     6.99   0.000     .0639343    .1137886
       timeGJ3 |   -.001385   .0002508    -5.52   0.000    -.0018765   -.0008935
         _cons |   -10.2459   1.590606    -6.44   0.000    -13.36343   -7.128367
---------------+----------------------------------------------------------------
      /lnsig2u |   3.030025   .2812598                      2.478766    3.581284
---------------+----------------------------------------------------------------
       sigma_u |   4.549479   .6397926                      3.453482      5.9933
           rho |   .8628514    .033284                      .7837947    .9160951
--------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 133.00                 Prob >= chibar2 = 0.000

. qui estimate store nullGJ

. xtlogit gjprog  W_SystemRELa persdumjlw_lag land timeGJ timeGJ2 timeGJ3, nolog

Random-effects logistic regression              Number of obs     =      5,335
Group variable: ccode                           Number of groups  =        173

Random effects u_i ~ Gaussian                   Obs per group:
                                                              min =          1
                                                              avg =       30.8
                                                              max =         55

Integration method: mvaghermite                 Integration pts.  =         12

                                                Wald chi2(6)      =     165.32
Log likelihood  = -191.66067                    Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
        gjprog |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
  W_SystemRELa |  -55.48516   15.56913    -3.56   0.000    -86.00008   -24.97023
persdumjlw_lag |    2.96594   .6461773     4.59   0.000     1.699456    4.232425
          land |    1.06616   .2302426     4.63   0.000     .6148932    1.517428
        timeGJ |  -1.659903   .1738081    -9.55   0.000    -2.000561   -1.319245
       timeGJ2 |   .0944066   .0128108     7.37   0.000     .0692979    .1195154
       timeGJ3 |  -.0014685     .00025    -5.87   0.000    -.0019586   -.0009785
         _cons |   2.096882    3.66717     0.57   0.567    -5.090639    9.284403
---------------+----------------------------------------------------------------
      /lnsig2u |   2.926887   .2851195                      2.368063    3.485711
---------------+----------------------------------------------------------------
       sigma_u |   4.320813   .6159741                      3.267521    5.713636
           rho |   .8501834   .0363161                      .7644468    .9084508
--------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 112.51                 Prob >= chibar2 = 0.000

. qui estimate store W_SystemRELaGJ

. xtlogit gjprog  reg5_w persdumjlw_lag land timeGJ timeGJ2 timeGJ3, nolog

Random-effects logistic regression              Number of obs     =      5,321
Group variable: ccode                           Number of groups  =        173

Random effects u_i ~ Gaussian                   Obs per group:
                                                              min =          1
                                                              avg =       30.8
                                                              max =         55

Integration method: mvaghermite                 Integration pts.  =         12

                                                Wald chi2(6)      =     163.23
Log likelihood  = -195.57235                    Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
        gjprog |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
        reg5_w |  -12.02935    5.45859    -2.20   0.028    -22.72799   -1.330714
persdumjlw_lag |   3.258122   .6774626     4.81   0.000      1.93032    4.585925
          land |   1.053278   .2370327     4.44   0.000     .5887024    1.517853
        timeGJ |  -1.516009   .1686218    -8.99   0.000    -1.846502   -1.185516
       timeGJ2 |   .0861715    .012709     6.78   0.000     .0612623    .1110806
       timeGJ3 |  -.0013418   .0002501    -5.37   0.000    -.0018319   -.0008516
         _cons |  -5.797986   2.399235    -2.42   0.016     -10.5004   -1.095571
---------------+----------------------------------------------------------------
      /lnsig2u |   3.168228    .266935                      2.645045    3.691411
---------------+----------------------------------------------------------------
       sigma_u |    4.87497   .6506499                      3.752876    6.332565
           rho |   .8784016   .0285119                      .8106436    .9241813
--------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 136.93                 Prob >= chibar2 = 0.000

. qui estimate store reg5_wGJ

. 
. esttab  nullSW W_SystemRELaSW reg5_wSW nullGJ W_SystemRELaGJ reg5_wGJ  , b(a2) se(2) replace label star(* 0.10 ** 0.05 *** 0.0
> 11) scalars(N chi2  ll ) pr2(4) varwidth(25) modelwidth(9)  ///
> title(Table 2: Uncertainty and Nuclear Proliferation)       ///
> nonumbers   ///
> mtitles("Model 1" "Model 2" "Model 3" "Model 4" "Model 5" "Model 6")  ///
> order(W_SystemRELa reg5_w persdumjlw_lag land) ///
> drop(timeSW timeSW2 timeSW3  timeGJ timeGJ2 timeGJ3 ) ///
> nogap

Table 2: Uncertainty and Nuclear Proliferation
-------------------------------------------------------------------------------------------------------
                            Model 1      Model 2      Model 3      Model 4      Model 5      Model 6   
-------------------------------------------------------------------------------------------------------
main                                                                                                   
W_SystemRELa                               -50.2***                               -55.5***             
                                         (14.63)                                (15.57)                
reg5_w                                                  -10.0*                                 -12.0** 
                                                       (5.59)                                 (5.46)   
persdumjlw_lag                 2.96***      2.77***      3.11***      3.30***      2.97***      3.26***
                             (0.64)       (0.62)       (0.68)       (0.65)       (0.65)       (0.68)   
land                           0.86***      0.95***      0.91***      1.06***      1.07***      1.05***
                             (0.20)       (0.19)       (0.23)       (0.23)       (0.23)       (0.24)   
Constant                      -10.3***     -0.72        -7.63***     -10.2***      2.10        -5.80** 
                             (1.50)       (3.32)       (2.23)       (1.59)       (3.67)       (2.40)   
-------------------------------------------------------------------------------------------------------
lnsig2u                                                                                                
Constant                       3.10***      3.31***      3.41***      3.03***      2.93***      3.17***
                             (0.27)       (0.22)       (0.30)       (0.28)       (0.29)       (0.27)   
-------------------------------------------------------------------------------------------------------
Observations                   5338         5338         5324         5337         5335         5321   
Pseudo R-squared                                                                                       
chi2                          143.7        145.9        139.5        162.8        165.3        163.2   
ll                           -210.9       -204.0       -208.6       -198.2       -191.7       -195.6   
-------------------------------------------------------------------------------------------------------
Standard errors in parentheses
* p<0.10, ** p<0.05, *** p<0.011

. 
. *************************************************************
. *Predicted Probabilities from Logit************************** -OK
. *************************************************************
. set seed 12345

. estsimp logit pursueonly W_SystemRELa persdumjlw_lag  land   timeSW timeSW2 timeSW3

Iteration 0:   log likelihood = -878.86806
Iteration 1:   log likelihood = -504.00465
Iteration 2:   log likelihood =  -348.8133
Iteration 3:   log likelihood =  -295.1144
Iteration 4:   log likelihood = -277.92812
Iteration 5:   log likelihood = -274.54419
Iteration 6:   log likelihood = -274.29285
Iteration 7:   log likelihood = -274.29005

Logistic regression                               Number of obs   =       5338
                                                  LR chi2(6)      =    1209.16
                                                  Prob > chi2     =     0.0000
Log likelihood = -274.29005                       Pseudo R2       =     0.6879

------------------------------------------------------------------------------
  pursueonly |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
W_SystemRELa |  -68.73278   11.50383    -5.97   0.000    -91.27988   -46.18568
persdumjlw~g |   2.275667   .2998004     7.59   0.000     1.688069    2.863265
        land |   .2879019   .0497447     5.79   0.000      .190404    .3853998
      timeSW |   -1.63569   .1201375   -13.62   0.000    -1.871155   -1.400225
     timeSW2 |   .0755331   .0074572    10.13   0.000     .0609172     .090149
     timeSW3 |  -.0009641   .0001244    -7.75   0.000    -.0012078   -.0007203
       _cons |   13.54617   2.410074     5.62   0.000     8.822511    18.26983
------------------------------------------------------------------------------
Note: 74 failures and 0 successes completely determined.

Simulating main parameters.  Please wait....
% of simulations completed: 14% 28% 42% 57% 71% 85% 100% 

Number of simulations  : 1000
Names of new variables : b1 b2 b3 b4 b5 b6 b7

. setx persdumjlw_lag 0 land 5 timeSW 0 timeSW2 0 timeSW3 0

. clarifyone W_SystemRELa 100
(7,414 missing values generated)


    Variable |        Obs        Mean    Std. Dev.       Min        Max
-------------+---------------------------------------------------------
xaxis_W_Sy~a |        100    .2142934    .0149612   .1887663   .2398204
(93 missing values generated)
(177 real changes made, 177 to missing)

. simqi, fd(prval(1)) changex(W_SystemRELa min max)

First Difference: W_SystemRELa min max

      Quantity of Interest |     Mean       Std. Err.    [95% Conf. Interval]
---------------------------+--------------------------------------------------
         dPr(pursue~y = 1) |  -.6959671     .0765953    -.8228305   -.5231245

. capture drop `e(allsims)'

. *********************************************
. ******SIMULATIONS**************************** -OK
. *********************************************
. ***************Pr(Pursuit|System Uncertainty)
. xtlogit pursueonly W_SystemRELa persdumjlw_lag  land   timeSW timeSW2 timeSW3, nolog

Random-effects logistic regression              Number of obs     =      5,338
Group variable: ccode                           Number of groups  =        173

Random effects u_i ~ Gaussian                   Obs per group:
                                                              min =          1
                                                              avg =       30.9
                                                              max =         55

Integration method: mvaghermite                 Integration pts.  =         12

                                                Wald chi2(6)      =     145.88
Log likelihood  = -204.00725                    Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
    pursueonly |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
  W_SystemRELa |  -50.17194   14.62579    -3.43   0.001    -78.83796   -21.50592
persdumjlw_lag |    2.76741   .6244856     4.43   0.000     1.543441    3.991379
          land |   .9482929   .1923574     4.93   0.000     .5712794    1.325306
        timeSW |  -1.221398    .120665   -10.12   0.000    -1.457897   -.9848987
       timeSW2 |   .0545218   .0068467     7.96   0.000     .0411025    .0679411
       timeSW3 |  -.0006356   .0001037    -6.13   0.000    -.0008388   -.0004325
         _cons |  -.7154484   3.320279    -0.22   0.829    -7.223075    5.792178
---------------+----------------------------------------------------------------
      /lnsig2u |   3.313633   .2245033                      2.873615    3.753652
---------------+----------------------------------------------------------------
       sigma_u |   5.242595   .5884899                      4.207243    6.532736
           rho |   .8930982   .0214342                      .8432706     .928429
--------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 140.57                 Prob >= chibar2 = 0.000

. matrix param=e(b)

. matrix P=e(V)

. save temp, replace
file temp.dta saved

. use temp, clear

. drawnorm bW_SystemRELa bpersdumjlw_lag  bland   btimeSW btimeSW2 btimeSW3 bcons blnsig2v, means(param) cov(P) double n(1000) s
> eed(12345)  clear
(obs 1,000)

. qui merge using temp

. expand 2 in 1
(1 observation created)

. 
. 
. qui replace persdumjlw_lag = 1 in 7515

. qui replace land = 10 in 7515

. qui replace timeSW = 0 in 7515

. qui replace timeSW2 = 0 in 7515

. qui replace timeSW3 = 0 in 7515

. 
. foreach x in 1 27{
  2. g p_`x'_Uncert=.
  3. forvalues i=1/1000 {
  4. qui replace W_SystemRELa =.188 +(`x'-1)*.002 in 7515
  5. qui generate Pr`x'_`i'=invlogit( bcons[`i']+bW_SystemRELa[`i']*W_SystemRELa[7515]+bpersdumjlw_lag[`i']*persdumjlw_lag[7515]
> + bland[`i']*land[7515]+ btimeSW[`i']*timeSW[7515]+ btimeSW2[`i']*timeSW2[7515]+ btimeSW3[`i']*timeSW3[7515])
  6. qui summarize Pr`x'_`i', meanonly
  7. qui replace p_`x'_Uncert = r(mean) in `i'
  8. drop Pr`x'_`i'
  9. }
 10. }
(7,515 missing values generated)
(7,515 missing values generated)

. g diff= p_27_Uncert- p_1_Uncert
(6,515 missing values generated)

. su  p_1_Uncert p_27_Uncert  diff 

    Variable |        Obs        Mean    Std. Dev.       Min        Max
-------------+---------------------------------------------------------
  p_1_Uncert |      1,000    .8304891    .1693907    .045599   .9980553
 p_27_Uncert |      1,000    .4125219    .2439739   .0064998   .9768268
        diff |      1,000   -.4179673    .1692351  -.7905415    .041386

. qui _pctile  p_1_Uncert, p(2.5 97.5)

. di r(r1) r(r2)
.36478071.99189001

. qui _pctile  p_27_Uncert, p(2.5 97.5)

. di r(r1) r(r2)
.0409996.88306057

. qui _pctile  diff, p(2.5 97.5)

. di r(r1) r(r2)
-.71460938-.09830502

. 
. drop _merge p_1_Uncert p_27_Uncert diff  bW_SystemRELa bpersdumjlw_lag bland btimeSW btimeSW2 btimeSW3 bcons blnsig2v

. 
. 
. ***************Pr(Pursuit|Fixed Region Uncertainty)
. 
. xtlogit pursueonly reg5_w persdumjlw_lag  land   timeSW timeSW2 timeSW3, nolog

Random-effects logistic regression              Number of obs     =      5,324
Group variable: ccode                           Number of groups  =        173

Random effects u_i ~ Gaussian                   Obs per group:
                                                              min =          1
                                                              avg =       30.8
                                                              max =         55

Integration method: mvaghermite                 Integration pts.  =         12

                                                Wald chi2(6)      =     139.48
Log likelihood  = -208.57261                    Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
    pursueonly |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
        reg5_w |  -10.01523   5.591433    -1.79   0.073    -20.97424    .9437758
persdumjlw_lag |   3.105664   .6830799     4.55   0.000     1.766852    4.444476
          land |    .906495   .2332235     3.89   0.000     .4493853    1.363605
        timeSW |  -1.144654   .1181002    -9.69   0.000    -1.376126   -.9131815
       timeSW2 |   .0515397   .0067596     7.62   0.000      .038291    .0647883
       timeSW3 |  -.0006048   .0001002    -6.04   0.000    -.0008012   -.0004085
         _cons |   -7.63486   2.227774    -3.43   0.001    -12.00122   -3.268502
---------------+----------------------------------------------------------------
      /lnsig2u |   3.410856   .3000548                       2.82276    3.998953
---------------+----------------------------------------------------------------
       sigma_u |   5.503742   .8257122                      4.101611    7.385189
           rho |    .902032   .0265159                      .8364314    .9431123
--------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 170.31                 Prob >= chibar2 = 0.000

. matrix param=e(b)

. matrix P=e(V)

. save temp, replace
file temp.dta saved

. use temp, clear

. drawnorm breg5_w bpersdumjlw_lag  bland   btimeSW btimeSW2 btimeSW3 bcons blnsig2v, means(param) cov(P) double n(1000) seed(12
> 345)  clear
(obs 1,000)

. qui merge using temp

. expand 2 in 1
(1 observation created)

. 
. 
. qui replace persdumjlw_lag = 1 in 7515

. qui replace land = 10 in 7515

. qui replace timeSW = 0 in 7515

. qui replace timeSW2 = 0 in 7515

. qui replace timeSW3 = 0 in 7515

. 
. foreach x in 1 18{
  2. g p_`x'_Uncert=.
  3. forvalues i=1/1000 {
  4. qui replace reg5_w =0.280 +(`x'-1)*.02  in 7515
  5. qui generate Pr`x'_`i'=invlogit( bcons[`i']+breg5_w[`i']*reg5_w[7515]+bpersdumjlw_lag[`i']*persdumjlw_lag[7515]+ bland[`i']
> *land[7515]+ btimeSW[`i']*timeSW[7515]+ btimeSW2[`i']*timeSW2[7515]+ btimeSW3[`i']*timeSW3[7515])
  6. qui summarize Pr`x'_`i', meanonly
  7. qui replace p_`x'_Uncert = r(mean) in `i'
  8. drop Pr`x'_`i'
  9. }
 10. }
(7,516 missing values generated)
(7,516 missing values generated)

. g diff= p_18_Uncert- p_1_Uncert
(6,516 missing values generated)

. su  p_1_Uncert p_18_Uncert  diff 

    Variable |        Obs        Mean    Std. Dev.       Min        Max
-------------+---------------------------------------------------------
  p_1_Uncert |      1,000    .7792727    .2039749   .0237163    .997409
 p_18_Uncert |      1,000    .2555843    .2449008   .0002201   .9953073
        diff |      1,000   -.5236884    .2775192  -.9740605   .6880063

. qui _pctile  p_1_Uncert, p(2.5 97.5)

. di r(r1) r(r2)
.23861717.99019378

. qui _pctile  p_18_Uncert, p(2.5 97.5)

. di r(r1) r(r2)
.00534162.85226202

. qui _pctile  diff, p(2.5 97.5)

. di r(r1) r(r2)
-.92986971.06639171

. 
. drop _merge  p_1_Uncert p_18_Uncert diff  breg5_w bpersdumjlw_lag bland btimeSW btimeSW2 btimeSW3 bcons blnsig2v

. 
. 
. 
. 
. 
. *********************************************
. **********TABLE 3**************************** - OK
. *********************************************
. *In-Sample and Out-of-Sample Performance - Nuclear Proliferation
. 
. g perrornullSW=.
(7,516 missing values generated)

. g perrorSysSW=.
(7,516 missing values generated)

. g perrorRegSW=.
(7,516 missing values generated)

. g perrornullGJ=.
(7,516 missing values generated)

. g perrorSysGJ=.
(7,516 missing values generated)

. g perrorRegGJ=.
(7,516 missing values generated)

. qui{

. 
. g cvlnullsw=-(pursueonly*log(0.5)+(1-pursueonly)*log(1-0.5))
(407 missing values generated)

. su cvlnullsw

    Variable |        Obs        Mean    Std. Dev.       Min        Max
-------------+---------------------------------------------------------
   cvlnullsw |      7,109    .6931472           0   .6931472   .6931472

. scalar nullswcvl=r(mean)

. 
. g cvlnullgj=-(gjprog*log(0.5)+(1-gjprog)*log(1-0.5))
(414 missing values generated)

. su cvlnullgj

    Variable |        Obs        Mean    Std. Dev.       Min        Max
-------------+---------------------------------------------------------
   cvlnullgj |      7,102    .6931472           0   .6931472   .6931472

. scalar nullgjcvl=r(mean)

. 
. 
. foreach v in perrornullSW perrorSysSW perrorRegSW {
  2. qui su `v'
  3. qui scalar PRI`v'=(.6931472-r(mean))/.6931472
  4. di  PRI`v'
  5. }
.63458971
.64550013
.55482233

. 
. 
. foreach v in  perrornullGJ perrorSysGJ perrorRegGJ{
  2. qui su `v'
  3. qui scalar PRI`v'=(.6931472-r(mean))/.6931472
  4. di  PRI`v'
  5. }
.68533136
.71919573
.62719513

. 
. 
. *********************************************
. *APPENDIX TABLE 4**************************** - OK 
. *********************************************
. *Europe and Africa as Reference Category
. eststo clear 

. xtlogit pursueonly W_SystemRELa persdumjlw_lag  land  Meast Asia Americas   timeSW timeSW2 timeSW3, nolog

Random-effects logistic regression              Number of obs     =      5,338
Group variable: ccode                           Number of groups  =        173

Random effects u_i ~ Gaussian                   Obs per group:
                                                              min =          1
                                                              avg =       30.9
                                                              max =         55

Integration method: mvaghermite                 Integration pts.  =         12

                                                Wald chi2(9)      =     151.41
Log likelihood  = -195.79744                    Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
    pursueonly |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
  W_SystemRELa |  -49.23138   14.61886    -3.37   0.001    -77.88381   -20.57895
persdumjlw_lag |   2.671036   .6406708     4.17   0.000     1.415345    3.926728
          land |   .9909508   .2164165     4.58   0.000     .5667822    1.415119
         Meast |   5.006725   1.673758     2.99   0.003     1.726219    8.287231
          Asia |    5.75373   1.651574     3.48   0.000     2.516705    8.990755
      Americas |   2.135226   1.889072     1.13   0.258    -1.567286    5.837739
        timeSW |   -1.23587   .1233532   -10.02   0.000    -1.477638   -.9941026
       timeSW2 |   .0558174   .0070774     7.89   0.000     .0419459    .0696889
       timeSW3 |  -.0006627   .0001076    -6.16   0.000    -.0008737   -.0004518
         _cons |  -2.311348   3.830912    -0.60   0.546    -9.819798    5.197101
---------------+----------------------------------------------------------------
      /lnsig2u |   2.816991   .3533636                      2.124411    3.509571
---------------+----------------------------------------------------------------
       sigma_u |   4.089798   .7225928                      2.892744    5.782208
           rho |   .8356407   .0485328                      .7177978    .9104159
--------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 81.03                  Prob >= chibar2 = 0.000

. qui estimate store W_SystemRELaSW

. xtlogit pursueonly reg5_w  persdumjlw_lag  land  Meast Asia Americas   timeSW timeSW2 timeSW3, nolog

Random-effects logistic regression              Number of obs     =      5,324
Group variable: ccode                           Number of groups  =        173

Random effects u_i ~ Gaussian                   Obs per group:
                                                              min =          1
                                                              avg =       30.8
                                                              max =         55

Integration method: mvaghermite                 Integration pts.  =         12

                                                Wald chi2(9)      =     155.63
Log likelihood  = -194.84534                    Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
    pursueonly |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
        reg5_w |  -25.80195   7.207135    -3.58   0.000    -39.92767   -11.67623
persdumjlw_lag |   3.238279   .7660915     4.23   0.000     1.736767     4.73979
          land |   .9789164   .2534422     3.86   0.000     .4821788    1.475654
         Meast |   11.90924   2.755079     4.32   0.000     6.509382    17.30909
          Asia |   7.210727   2.054893     3.51   0.000     3.183211    11.23824
      Americas |   .1115358   2.519768     0.04   0.965    -4.827119     5.05019
        timeSW |  -1.163039   .1221442    -9.52   0.000    -1.402437   -.9236404
       timeSW2 |   .0527883   .0070345     7.50   0.000     .0390009    .0665757
       timeSW3 |  -.0006267   .0001038    -6.04   0.000    -.0008302   -.0004232
         _cons |  -5.531321    4.12439    -1.34   0.180    -13.61498    2.552335
---------------+----------------------------------------------------------------
      /lnsig2u |   3.261605   .4777092                      2.325312    4.197898
---------------+----------------------------------------------------------------
       sigma_u |   5.107972   1.220063                      3.198417    8.157591
           rho |   .8880283   .0475006                      .7566619    .9528916
--------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 94.16                  Prob >= chibar2 = 0.000

. qui estimate store reg5_wSW

. xtlogit gjprog  W_SystemRELa persdumjlw_lag land  Meast Asia Americas  timeGJ timeGJ2 timeGJ3, nolog

Random-effects logistic regression              Number of obs     =      5,335
Group variable: ccode                           Number of groups  =        173

Random effects u_i ~ Gaussian                   Obs per group:
                                                              min =          1
                                                              avg =       30.8
                                                              max =         55

Integration method: mvaghermite                 Integration pts.  =         12

                                                Wald chi2(9)      =     166.99
Log likelihood  = -188.05007                    Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
        gjprog |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
  W_SystemRELa |  -53.91703   15.58603    -3.46   0.001    -84.46509   -23.36898
persdumjlw_lag |   2.904281   .6571094     4.42   0.000      1.61637    4.192191
          land |   1.071429   .2403453     4.46   0.000     .6003604    1.542497
         Meast |   2.796621   1.446406     1.93   0.053    -.0382839    5.631525
          Asia |   3.559784    1.36929     2.60   0.009     .8760242    6.243543
      Americas |   .6879176   1.834862     0.37   0.708    -2.908345     4.28418
        timeGJ |  -1.677644   .1762994    -9.52   0.000    -2.023184   -1.332103
       timeGJ2 |   .0957481   .0130251     7.35   0.000     .0702195    .1212768
       timeGJ3 |  -.0014939   .0002544    -5.87   0.000    -.0019925   -.0009953
         _cons |   .9175326   3.882877     0.24   0.813    -6.692766    8.527831
---------------+----------------------------------------------------------------
      /lnsig2u |   2.725555   .3461239                      2.047164    3.403945
---------------+----------------------------------------------------------------
       sigma_u |   3.907029   .6761581                      2.783146    5.484755
           rho |    .822694   .0504886                      .7018905    .9014195
--------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 103.52                 Prob >= chibar2 = 0.000

. qui estimate store W_SystemRELaGJ

. xtlogit gjprog  reg5_w persdumjlw_lag land  Meast Asia Americas  timeGJ timeGJ2 timeGJ3, nolog

Random-effects logistic regression              Number of obs     =      5,321
Group variable: ccode                           Number of groups  =        173

Random effects u_i ~ Gaussian                   Obs per group:
                                                              min =          1
                                                              avg =       30.8
                                                              max =         55

Integration method: mvaghermite                 Integration pts.  =         12

                                                Wald chi2(9)      =     170.08
Log likelihood  = -186.53504                    Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
        gjprog |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
        reg5_w |  -28.33922   7.230519    -3.92   0.000    -42.51078   -14.16767
persdumjlw_lag |   3.161598   .7202346     4.39   0.000     1.749965    4.573232
          land |    .848097   .2348975     3.61   0.000     .3877064    1.308488
         Meast |   8.614165   2.031137     4.24   0.000     4.633211    12.59512
          Asia |   3.573223   1.352723     2.64   0.008     .9219344    6.224512
      Americas |  -1.415698   1.804093    -0.78   0.433    -4.951655     2.12026
        timeGJ |  -1.528033    .170228    -8.98   0.000    -1.861674   -1.194393
       timeGJ2 |   .0862724   .0128144     6.73   0.000     .0611567    .1113882
       timeGJ3 |   -.001344    .000251    -5.35   0.000    -.0018359   -.0008521
         _cons |   1.480733   3.601923     0.41   0.681    -5.578907    8.540373
---------------+----------------------------------------------------------------
      /lnsig2u |   2.496019   .4290475                      1.655102    3.336937
---------------+----------------------------------------------------------------
       sigma_u |   3.483403   .7472725                      2.287709    5.304038
           rho |   .7867041   .0719945                      .6140229    .8953028
--------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 78.90                  Prob >= chibar2 = 0.000

. qui estimate store reg5_wGJ

. 
. esttab  W_SystemRELaSW reg5_wSW W_SystemRELaGJ reg5_wGJ, b(a2) se(2) replace label star(* 0.10 ** 0.05 *** 0.011) scalars(N ch
> i2  ll ) pr2(4) varwidth(25) modelwidth(9)  ///
> title(Table A4: Uncertainty and Nuclear Proliferation)       ///
> nonumbers   ///
> mtitles("Model 1" "Model 2" "Model 3" "Model 4" "Model 5" "Model 6")  ///
> order(W_SystemRELa reg5_w persdumjlw_lag land) ///
> drop(timeSW timeSW2 timeSW3  timeGJ timeGJ2 timeGJ3 ) ///
> nogap

Table A4: Uncertainty and Nuclear Proliferation
-----------------------------------------------------------------------------
                            Model 1      Model 2      Model 3      Model 4   
-----------------------------------------------------------------------------
main                                                                         
W_SystemRELa                  -49.2***                  -53.9***             
                            (14.62)                   (15.59)                
reg5_w                                     -25.8***                  -28.3***
                                          (7.21)                    (7.23)   
persdumjlw_lag                 2.67***      3.24***      2.90***      3.16***
                             (0.64)       (0.77)       (0.66)       (0.72)   
land                           0.99***      0.98***      1.07***      0.85***
                             (0.22)       (0.25)       (0.24)       (0.23)   
Middle East                    5.01***      11.9***      2.80*        8.61***
                             (1.67)       (2.76)       (1.45)       (2.03)   
Asia                           5.75***      7.21***      3.56***      3.57***
                             (1.65)       (2.05)       (1.37)       (1.35)   
America                        2.14         0.11         0.69        -1.42   
                             (1.89)       (2.52)       (1.83)       (1.80)   
Constant                      -2.31        -5.53         0.92         1.48   
                             (3.83)       (4.12)       (3.88)       (3.60)   
-----------------------------------------------------------------------------
lnsig2u                                                                      
Constant                       2.82***      3.26***      2.73***      2.50***
                             (0.35)       (0.48)       (0.35)       (0.43)   
-----------------------------------------------------------------------------
Observations                   5338         5324         5335         5321   
Pseudo R-squared                                                             
chi2                          151.4        155.6        167.0        170.1   
ll                           -195.8       -194.8       -188.1       -186.5   
-----------------------------------------------------------------------------
Standard errors in parentheses
* p<0.10, ** p<0.05, *** p<0.011

. 
. 
. 
. *********************************************
. *APPENDIX TABLE 7**************************** - OK 
. *********************************************
. *Uncertainty and Nuclear Proliferation Singh and Way | DOE Scores
. 
. eststo clear

. foreach v in SystemE W_SystemE W_System_d1E W_System_d2E W_System_d1cE W_System_d2cE {
  2. xtlogit pursueonly  `v'    persdumjlw_lag  land   timeSW timeSW2 timeSW3, nolog
  3. qui estimate store _est`v'
  4. }

Random-effects logistic regression              Number of obs     =      5,338
Group variable: ccode                           Number of groups  =        173

Random effects u_i ~ Gaussian                   Obs per group:
                                                              min =          1
                                                              avg =       30.9
                                                              max =         55

Integration method: mvaghermite                 Integration pts.  =         12

                                                Wald chi2(6)      =     144.50
Log likelihood  = -201.85284                    Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
    pursueonly |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
       SystemE |  -21.66939   5.623867    -3.85   0.000    -32.69197   -10.64682
persdumjlw_lag |   3.105177   .7186492     4.32   0.000      1.69665    4.513703
          land |   .7275537   .2522334     2.88   0.004     .2331853    1.221922
        timeSW |  -1.061208   .1152458    -9.21   0.000    -1.287085     -.83533
       timeSW2 |   .0465859   .0067052     6.95   0.000      .033444    .0597278
       timeSW3 |  -.0005451   .0001001    -5.45   0.000    -.0007412    -.000349
         _cons |  -2.999572   2.648663    -1.13   0.257    -8.190857    2.191713
---------------+----------------------------------------------------------------
      /lnsig2u |   3.312207   .3027334                       2.71886    3.905553
---------------+----------------------------------------------------------------
       sigma_u |   5.238857   .7929883                      3.893973     7.04823
           rho |   .8929619   .0289355                      .8217153    .9378888
--------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 155.27                 Prob >= chibar2 = 0.000

Random-effects logistic regression              Number of obs     =      5,338
Group variable: ccode                           Number of groups  =        173

Random effects u_i ~ Gaussian                   Obs per group:
                                                              min =          1
                                                              avg =       30.9
                                                              max =         55

Integration method: mvaghermite                 Integration pts.  =         12

                                                Wald chi2(6)      =     153.96
Log likelihood  = -198.78057                    Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
    pursueonly |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
     W_SystemE |  -28.01642   6.684199    -4.19   0.000    -41.11721   -14.91564
persdumjlw_lag |   3.010884   .6775652     4.44   0.000     1.682881    4.338888
          land |   .7120072   .1983919     3.59   0.000     .3231661    1.100848
        timeSW |  -1.054384   .1118057    -9.43   0.000    -1.273519   -.8352491
       timeSW2 |   .0459057   .0064279     7.14   0.000     .0333073    .0585041
       timeSW3 |  -.0005332   .0000958    -5.57   0.000    -.0007209   -.0003455
         _cons |   .9994828   2.910106     0.34   0.731    -4.704219    6.703185
---------------+----------------------------------------------------------------
      /lnsig2u |   3.265455   .2291073                      2.816413    3.714497
---------------+----------------------------------------------------------------
       sigma_u |   5.117815   .5862643                      4.088617    6.406087
           rho |   .8884106   .0227131                      .8355613    .9257832
--------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 155.77                 Prob >= chibar2 = 0.000

Random-effects logistic regression              Number of obs     =      5,338
Group variable: ccode                           Number of groups  =        173

Random effects u_i ~ Gaussian                   Obs per group:
                                                              min =          1
                                                              avg =       30.9
                                                              max =         55

Integration method: mvaghermite                 Integration pts.  =         12

                                                Wald chi2(6)      =     145.95
Log likelihood  = -202.55873                    Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
    pursueonly |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
  W_System_d1E |  -18.56707   5.021209    -3.70   0.000    -28.40846   -8.725679
persdumjlw_lag |   3.096641   .7040272     4.40   0.000     1.716773    4.476509
          land |   .7337388   .2340023     3.14   0.002     .2751028    1.192375
        timeSW |  -1.065285   .1144424    -9.31   0.000    -1.289587   -.8409816
       timeSW2 |   .0469481   .0066528     7.06   0.000     .0339089    .0599873
       timeSW3 |  -.0005502   .0000996    -5.52   0.000    -.0007454    -.000355
         _cons |   -4.44716   2.387669    -1.86   0.063    -9.126904    .2325846
---------------+----------------------------------------------------------------
      /lnsig2u |   3.383982   .2652378                      2.864126    3.903839
---------------+----------------------------------------------------------------
       sigma_u |   5.430282   .7201581                      4.187328    7.042192
           rho |   .8996313   .0239496                      .8420124    .9377889
--------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 157.38                 Prob >= chibar2 = 0.000

Random-effects logistic regression              Number of obs     =      5,338
Group variable: ccode                           Number of groups  =        173

Random effects u_i ~ Gaussian                   Obs per group:
                                                              min =          1
                                                              avg =       30.9
                                                              max =         55

Integration method: mvaghermite                 Integration pts.  =         12

                                                Wald chi2(6)      =     145.94
Log likelihood  = -202.54906                    Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
    pursueonly |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
  W_System_d2E |  -18.56529   5.017753    -3.70   0.000     -28.3999   -8.730672
persdumjlw_lag |   3.096761   .7041347     4.40   0.000     1.716682     4.47684
          land |   .7336153   .2341328     3.13   0.002     .2747235    1.192507
        timeSW |  -1.065121   .1144473    -9.31   0.000    -1.289434   -.8408085
       timeSW2 |   .0469372    .006653     7.06   0.000     .0338976    .0599769
       timeSW3 |  -.0005501   .0000996    -5.52   0.000    -.0007453   -.0003549
         _cons |   -4.44635   2.387448    -1.86   0.063    -9.125663     .232963
---------------+----------------------------------------------------------------
      /lnsig2u |   3.383466   .2654622                      2.863169    3.903762
---------------+----------------------------------------------------------------
       sigma_u |    5.42888   .7205812                      4.185326    7.041921
           rho |   .8995847   .0239798                      .8418851    .9377844
--------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 157.35                 Prob >= chibar2 = 0.000

Random-effects logistic regression              Number of obs     =      5,338
Group variable: ccode                           Number of groups  =        173

Random effects u_i ~ Gaussian                   Obs per group:
                                                              min =          1
                                                              avg =       30.9
                                                              max =         55

Integration method: mvaghermite                 Integration pts.  =         12

                                                Wald chi2(6)      =     147.20
Log likelihood  =  -197.3216                    Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
    pursueonly |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
 W_System_d1cE |  -26.48764   6.010979    -4.41   0.000    -38.26895   -14.70634
persdumjlw_lag |   3.112682   .7285008     4.27   0.000     1.684847    4.540518
          land |   .7329322   .2241414     3.27   0.001     .2936231    1.172241
        timeSW |  -1.038572   .1136359    -9.14   0.000    -1.261294   -.8158495
       timeSW2 |   .0448892   .0064967     6.91   0.000     .0321559    .0576225
       timeSW3 |  -.0005136   .0000955    -5.38   0.000    -.0007009   -.0003264
         _cons |  -1.724441   2.798149    -0.62   0.538    -7.208712    3.759829
---------------+----------------------------------------------------------------
      /lnsig2u |   3.728268   .2182946                      3.300418    4.156117
---------------+----------------------------------------------------------------
       sigma_u |   6.450347   .7040379                      5.208069    7.988945
           rho |   .9267239   .0148237                        .89183    .9509802
--------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 158.01                 Prob >= chibar2 = 0.000

Random-effects logistic regression              Number of obs     =      5,338
Group variable: ccode                           Number of groups  =        173

Random effects u_i ~ Gaussian                   Obs per group:
                                                              min =          1
                                                              avg =       30.9
                                                              max =         55

Integration method: mvaghermite                 Integration pts.  =         12

                                                Wald chi2(6)      =     148.11
Log likelihood  = -197.43699                    Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
    pursueonly |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
 W_System_d2cE |    -25.917   5.896416    -4.40   0.000    -37.47376   -14.36024
persdumjlw_lag |   3.099246    .717689     4.32   0.000     1.692601    4.505891
          land |   .7104126   .2213164     3.21   0.001     .2766405    1.144185
        timeSW |   -1.03403    .113122    -9.14   0.000    -1.255745   -.8123147
       timeSW2 |   .0446096   .0064851     6.88   0.000     .0318991    .0573201
       timeSW3 |   -.000512   .0000957    -5.35   0.000    -.0006997   -.0003244
         _cons |  -.9025093   2.742146    -0.33   0.742    -6.277017    4.471999
---------------+----------------------------------------------------------------
      /lnsig2u |   3.526381   .2293804                      3.076804    3.975958
---------------+----------------------------------------------------------------
       sigma_u |   5.831012   .6687599                      4.657142    7.300766
           rho |   .9117775   .0184512                      .8682938    .9418659
--------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 157.24                 Prob >= chibar2 = 0.000

. 
. esttab  _estSystemE _estW_SystemE _estW_System_d1E _estW_System_d2E _estW_System_d1cE _estW_System_d2cE, b(a2) se(2) replace s
> tar(* 0.10 ** 0.05 *** 0.011) scalars(N chi2  ll ) pr2(4) varwidth(25) modelwidth(9)  ///
> title(Table A7: Uncertainty and Nuclear Proliferation)       ///
> nonumbers   ///
> mtitles("Model 1" "Model 2" "Model 3" "Model 4" "Model 5" "Model 6")  ///
> order(SystemE W_SystemE W_System_d1E W_System_d2E W_System_d1cE W_System_d2cE) ///
> drop(timeSW timeSW2 timeSW3   ) ///
> nogap

Table A7: Uncertainty and Nuclear Proliferation
-------------------------------------------------------------------------------------------------------
                            Model 1      Model 2      Model 3      Model 4      Model 5      Model 6   
-------------------------------------------------------------------------------------------------------
pursueonly                                                                                             
SystemE                       -21.7***                                                                 
                             (5.62)                                                                    
W_SystemE                                  -28.0***                                                    
                                          (6.68)                                                       
W_System_d1E                                            -18.6***                                       
                                                       (5.02)                                          
W_System_d2E                                                         -18.6***                          
                                                                    (5.02)                             
W_System_d1cE                                                                     -26.5***             
                                                                                 (6.01)                
W_System_d2cE                                                                                  -25.9***
                                                                                              (5.90)   
persdumjlw_lag                 3.11***      3.01***      3.10***      3.10***      3.11***      3.10***
                             (0.72)       (0.68)       (0.70)       (0.70)       (0.73)       (0.72)   
land                           0.73***      0.71***      0.73***      0.73***      0.73***      0.71***
                             (0.25)       (0.20)       (0.23)       (0.23)       (0.22)       (0.22)   
_cons                         -3.00         1.00        -4.45*       -4.45*       -1.72        -0.90   
                             (2.65)       (2.91)       (2.39)       (2.39)       (2.80)       (2.74)   
-------------------------------------------------------------------------------------------------------
lnsig2u                                                                                                
_cons                          3.31***      3.27***      3.38***      3.38***      3.73***      3.53***
                             (0.30)       (0.23)       (0.27)       (0.27)       (0.22)       (0.23)   
-------------------------------------------------------------------------------------------------------
N                              5338         5338         5338         5338         5338         5338   
pseudo R-sq                                                                                            
chi2                          144.5        154.0        145.9        145.9        147.2        148.1   
ll                           -201.9       -198.8       -202.6       -202.5       -197.3       -197.4   
-------------------------------------------------------------------------------------------------------
Standard errors in parentheses
* p<0.10, ** p<0.05, *** p<0.011

. 
. 
. drop *est_*

. *********************************************
. *APPENDIX TABLE 8**************************** - OK
. *********************************************
. *Uncertainty and Nuclear Proliferation Gartzke and Jo | DOE Scores
. foreach v in SystemE W_SystemE W_System_d1E W_System_d2E W_System_d1cE W_System_d2cE{
  2. xtlogit gjprog `v'   persdumjlw_lag land timeGJ timeGJ2 timeGJ3, nolog
  3. qui estimate store _est`v'
  4. }

Random-effects logistic regression              Number of obs     =      5,337
Group variable: ccode                           Number of groups  =        173

Random effects u_i ~ Gaussian                   Obs per group:
                                                              min =          1
                                                              avg =       30.8
                                                              max =         55

Integration method: mvaghermite                 Integration pts.  =         12

                                                Wald chi2(6)      =     164.87
Log likelihood  = -192.59979                    Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
        gjprog |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
       SystemE |  -16.10048     4.9271    -3.27   0.001    -25.75742   -6.443542
persdumjlw_lag |   3.313245   .6885887     4.81   0.000     1.963636    4.662854
          land |   .8504096   .2281314     3.73   0.000     .4032802    1.297539
        timeGJ |  -1.458168   .1655738    -8.81   0.000    -1.782687   -1.133649
       timeGJ2 |   .0820747   .0125056     6.56   0.000     .0575642    .1065851
       timeGJ3 |  -.0012851   .0002467    -5.21   0.000    -.0017687   -.0008015
         _cons |  -3.633048   2.382494    -1.52   0.127    -8.302651    1.036555
---------------+----------------------------------------------------------------
      /lnsig2u |   3.023538   .2635771                      2.506936     3.54014
---------------+----------------------------------------------------------------
       sigma_u |   4.534746   .5976276                      3.502469    5.871263
           rho |   .8620819   .0313384                      .7885302    .9128779
--------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 128.38                 Prob >= chibar2 = 0.000

Random-effects logistic regression              Number of obs     =      5,337
Group variable: ccode                           Number of groups  =        173

Random effects u_i ~ Gaussian                   Obs per group:
                                                              min =          1
                                                              avg =       30.8
                                                              max =         55

Integration method: mvaghermite                 Integration pts.  =         12

                                                Wald chi2(6)      =     163.87
Log likelihood  = -191.34626                    Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
        gjprog |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
     W_SystemE |  -19.98156   5.719677    -3.49   0.000    -31.19193   -8.771204
persdumjlw_lag |   3.275696   .6928436     4.73   0.000     1.917747    4.633645
          land |   .8466534   .2329068     3.64   0.000     .3901645    1.303142
        timeGJ |  -1.460744   .1652431    -8.84   0.000    -1.784615   -1.136873
       timeGJ2 |   .0818482   .0124426     6.58   0.000     .0574612    .1062352
       timeGJ3 |  -.0012768   .0002451    -5.21   0.000    -.0017572   -.0007964
         _cons |   -1.42127    2.82824    -0.50   0.615    -6.964518    4.121978
---------------+----------------------------------------------------------------
      /lnsig2u |   3.077374    .254852                      2.577873    3.576875
---------------+----------------------------------------------------------------
       sigma_u |    4.65847   .5936102                      3.628925    5.980101
           rho |    .868359   .0291326                       .800117    .9157556
--------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 127.04                 Prob >= chibar2 = 0.000

Random-effects logistic regression              Number of obs     =      5,337
Group variable: ccode                           Number of groups  =        173

Random effects u_i ~ Gaussian                   Obs per group:
                                                              min =          1
                                                              avg =       30.8
                                                              max =         55

Integration method: mvaghermite                 Integration pts.  =         12

                                                Wald chi2(6)      =     164.00
Log likelihood  = -193.18516                    Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
        gjprog |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
  W_System_d1E |  -13.97791   4.458653    -3.14   0.002    -22.71671   -5.239111
persdumjlw_lag |   3.320637   .6908206     4.81   0.000     1.966653     4.67462
          land |   .8574069   .2320514     3.69   0.000     .4025946    1.312219
        timeGJ |  -1.459492   .1658136    -8.80   0.000     -1.78448   -1.134503
       timeGJ2 |   .0822268   .0125184     6.57   0.000     .0576912    .1067625
       timeGJ3 |  -.0012883   .0002468    -5.22   0.000    -.0017721   -.0008045
         _cons |  -4.366085   2.295878    -1.90   0.057    -8.865923    .1337525
---------------+----------------------------------------------------------------
      /lnsig2u |   3.007445   .2732768                      2.471833    3.543058
---------------+----------------------------------------------------------------
       sigma_u |   4.498404   .6146547                      3.441531    5.879837
           rho |   .8601574   .0328715                      .7826174    .9131098
--------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 129.48                 Prob >= chibar2 = 0.000

Random-effects logistic regression              Number of obs     =      5,337
Group variable: ccode                           Number of groups  =        173

Random effects u_i ~ Gaussian                   Obs per group:
                                                              min =          1
                                                              avg =       30.8
                                                              max =         55

Integration method: mvaghermite                 Integration pts.  =         12

                                                Wald chi2(6)      =     163.94
Log likelihood  = -193.20535                    Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
        gjprog |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
  W_System_d2E |  -13.95215   4.453481    -3.13   0.002    -22.68081   -5.223482
persdumjlw_lag |    3.32025   .6912671     4.80   0.000     1.965391    4.675108
          land |   .8572592   .2325561     3.69   0.000     .4014576    1.313061
        timeGJ |  -1.459613   .1658286    -8.80   0.000    -1.784631   -1.134595
       timeGJ2 |   .0822355   .0125195     6.57   0.000     .0576976    .1067733
       timeGJ3 |  -.0012885   .0002469    -5.22   0.000    -.0017724   -.0008047
         _cons |  -4.363273   2.299249    -1.90   0.058    -8.869719    .1431725
---------------+----------------------------------------------------------------
      /lnsig2u |   3.003955   .2749028                      2.465155    3.542755
---------------+----------------------------------------------------------------
       sigma_u |   4.490561   .6172339                       3.43006    5.878945
           rho |    .859737   .0331503                      .7814793    .9130857
--------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 129.43                 Prob >= chibar2 = 0.000

Random-effects logistic regression              Number of obs     =      5,337
Group variable: ccode                           Number of groups  =        173

Random effects u_i ~ Gaussian                   Obs per group:
                                                              min =          1
                                                              avg =       30.8
                                                              max =         55

Integration method: mvaghermite                 Integration pts.  =         12

                                                Wald chi2(6)      =     163.71
Log likelihood  = -192.14015                    Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
        gjprog |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
 W_System_d1cE |  -16.16557   4.854191    -3.33   0.001    -25.67961   -6.651534
persdumjlw_lag |   3.291507   .6936001     4.75   0.000     1.932076    4.650938
          land |   .8579122   .2319914     3.70   0.000     .4032173    1.312607
        timeGJ |   -1.45217   .1657193    -8.76   0.000    -1.776974   -1.127366
       timeGJ2 |    .081196   .0124768     6.51   0.000     .0567419    .1056502
       timeGJ3 |  -.0012642   .0002454    -5.15   0.000    -.0017452   -.0007831
         _cons |  -2.850935   2.583052    -1.10   0.270    -7.913624    2.211754
---------------+----------------------------------------------------------------
      /lnsig2u |   3.068067   .2558579                      2.566594    3.569539
---------------+----------------------------------------------------------------
       sigma_u |   4.636841   .5931863                      3.608518    5.958206
           rho |   .8672914   .0294485                      .7983071    .9151879
--------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 127.64                 Prob >= chibar2 = 0.000

Random-effects logistic regression              Number of obs     =      5,337
Group variable: ccode                           Number of groups  =        173

Random effects u_i ~ Gaussian                   Obs per group:
                                                              min =          1
                                                              avg =       30.8
                                                              max =         55

Integration method: mvaghermite                 Integration pts.  =         12

                                                Wald chi2(6)      =     163.72
Log likelihood  = -192.16763                    Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
        gjprog |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
 W_System_d2cE |   -16.0118   4.805821    -3.33   0.001    -25.43104   -6.592568
persdumjlw_lag |   3.290594   .6942404     4.74   0.000     1.929908     4.65128
          land |   .8573129   .2326362     3.69   0.000     .4013543    1.313271
        timeGJ |  -1.452872   .1657141    -8.77   0.000    -1.777666   -1.128078
       timeGJ2 |    .081233    .012476     6.51   0.000     .0567804    .1056856
       timeGJ3 |  -.0012648   .0002454    -5.15   0.000    -.0017459   -.0007838
         _cons |  -2.890632   2.566759    -1.13   0.260    -7.921387    2.140123
---------------+----------------------------------------------------------------
      /lnsig2u |   3.054656   .2591412                      2.546748    3.562563
---------------+----------------------------------------------------------------
       sigma_u |   4.605853   .5967832                      3.572888    5.937461
           rho |   .8657402    .030121                      .7950927    .9146449
--------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 127.35                 Prob >= chibar2 = 0.000

. 
. 
. esttab    _estSystemE _estW_SystemE _estW_System_d1E _estW_System_d2E _estW_System_d1cE _estW_System_d2cE, b(a2) se(2) replace
>  star(* 0.10 ** 0.05 *** 0.011) scalars(N chi2  ll ) pr2(4) varwidth(25) modelwidth(9)  ///
> title(Table 2: Uncertainty and Nuclear Proliferation)       ///
> nonumbers   ///
> mtitles("Model 1" "Model 2" "Model 3" "Model 4" "Model 5" "Model 6")  ///
> order(SystemE W_SystemE W_System_d1E W_System_d2E W_System_d1cE W_System_d2cE ) ///
> drop(time* ) ///
> nogap

Table 2: Uncertainty and Nuclear Proliferation
-------------------------------------------------------------------------------------------------------
                            Model 1      Model 2      Model 3      Model 4      Model 5      Model 6   
-------------------------------------------------------------------------------------------------------
gjprog                                                                                                 
SystemE                       -16.1***                                                                 
                             (4.93)                                                                    
W_SystemE                                  -20.0***                                                    
                                          (5.72)                                                       
W_System_d1E                                            -14.0***                                       
                                                       (4.46)                                          
W_System_d2E                                                         -14.0***                          
                                                                    (4.45)                             
W_System_d1cE                                                                     -16.2***             
                                                                                 (4.85)                
W_System_d2cE                                                                                  -16.0***
                                                                                              (4.81)   
persdumjlw_lag                 3.31***      3.28***      3.32***      3.32***      3.29***      3.29***
                             (0.69)       (0.69)       (0.69)       (0.69)       (0.69)       (0.69)   
land                           0.85***      0.85***      0.86***      0.86***      0.86***      0.86***
                             (0.23)       (0.23)       (0.23)       (0.23)       (0.23)       (0.23)   
_cons                         -3.63        -1.42        -4.37*       -4.36*       -2.85        -2.89   
                             (2.38)       (2.83)       (2.30)       (2.30)       (2.58)       (2.57)   
-------------------------------------------------------------------------------------------------------
lnsig2u                                                                                                
_cons                          3.02***      3.08***      3.01***      3.00***      3.07***      3.05***
                             (0.26)       (0.25)       (0.27)       (0.27)       (0.26)       (0.26)   
-------------------------------------------------------------------------------------------------------
N                              5337         5337         5337         5337         5337         5337   
pseudo R-sq                                                                                            
chi2                          164.9        163.9        164.0        163.9        163.7        163.7   
ll                           -192.6       -191.3       -193.2       -193.2       -192.1       -192.2   
-------------------------------------------------------------------------------------------------------
Standard errors in parentheses
* p<0.10, ** p<0.05, *** p<0.011

. 
. 
. 
. *********************************************
. ******APPENDIX TABLE 10********************** - OK
. *********************************************
. *Uncertainty and Nuclear Proliferation, Singh and Way | Hierarchical Model
. btscs pursueonly  year ccode, g(swt) nspline(3)

. 
. set more off

. eststo:  xtmelogit pursueonly              persdumjlw_lag  land   swt _spline1 _spline2 _spline3  || year:, intpoints(10)

Refining starting values: 

Iteration 0:   log likelihood = -238.02916  (not concave)
Iteration 1:   log likelihood = -234.95758  
Iteration 2:   log likelihood = -232.87161  

Performing gradient-based optimization: 

Iteration 0:   log likelihood = -232.87161  
Iteration 1:   log likelihood =  -232.8477  
Iteration 2:   log likelihood = -232.84391  
Iteration 3:   log likelihood = -232.84391  

Mixed-effects logistic regression               Number of obs     =      5,338
Group variable: year                            Number of groups  =         55

                                                Obs per group:
                                                              min =         31
                                                              avg =       97.1
                                                              max =        141

Integration points =  10                        Wald chi2(6)      =     326.32
Log likelihood = -232.84391                     Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
    pursueonly |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
persdumjlw_lag |   1.850674   .3558159     5.20   0.000     1.153287     2.54806
          land |   .2451585   .0601667     4.07   0.000     .1272339    .3630831
           swt |  -2.634566    .250342   -10.52   0.000    -3.125227   -2.143905
      _spline1 |  -.0219337   .0024952    -8.79   0.000    -.0268241   -.0170433
      _spline2 |   .0132347    .001676     7.90   0.000     .0099498    .0165197
      _spline3 |  -.0032625   .0005899    -5.53   0.000    -.0044188   -.0021063
         _cons |   .0277855   .3271599     0.08   0.932    -.6134362    .6690072
--------------------------------------------------------------------------------

------------------------------------------------------------------------------
  Random-effects Parameters  |   Estimate   Std. Err.     [95% Conf. Interval]
-----------------------------+------------------------------------------------
year: Identity               |
                   sd(_cons) |   .9900508   .1850217      .6864108    1.428009
------------------------------------------------------------------------------
LR test vs. logistic model: chibar2(01) = 40.16       Prob >= chibar2 = 0.0000
(est1 stored)

. eststo:  xtmelogit pursueonly W_SystemRELa persdumjlw_lag  land   swt _spline1 _spline2 _spline3  || year:, intpoints(10)

Refining starting values: 

Iteration 0:   log likelihood = -230.54884  (not concave)
Iteration 1:   log likelihood = -227.35478  (not concave)
Iteration 2:   log likelihood = -226.88628  

Performing gradient-based optimization: 

Iteration 0:   log likelihood = -226.88628  
Iteration 1:   log likelihood = -225.94737  
Iteration 2:   log likelihood = -225.91113  
Iteration 3:   log likelihood = -225.91073  
Iteration 4:   log likelihood = -225.91073  

Mixed-effects logistic regression               Number of obs     =      5,338
Group variable: year                            Number of groups  =         55

                                                Obs per group:
                                                              min =         31
                                                              avg =       97.1
                                                              max =        141

Integration points =  10                        Wald chi2(7)      =     323.17
Log likelihood = -225.91073                     Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
    pursueonly |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
  W_SystemRELa |  -66.53969   17.19186    -3.87   0.000    -100.2351   -32.84427
persdumjlw_lag |     2.0219   .3646764     5.54   0.000     1.307148    2.736653
          land |   .2640876   .0596181     4.43   0.000     .1472381     .380937
           swt |   -2.54916   .2440445   -10.45   0.000    -3.027479   -2.070842
      _spline1 |  -.0208319   .0024174    -8.62   0.000    -.0255698   -.0160939
      _spline2 |   .0124326   .0016171     7.69   0.000     .0092631    .0156021
      _spline3 |  -.0029943   .0005656    -5.29   0.000    -.0041027   -.0018858
         _cons |    13.9512   3.612676     3.86   0.000     6.870489    21.03192
--------------------------------------------------------------------------------

------------------------------------------------------------------------------
  Random-effects Parameters  |   Estimate   Std. Err.     [95% Conf. Interval]
-----------------------------+------------------------------------------------
year: Identity               |
                   sd(_cons) |   .7663047   .1653455      .5020394    1.169675
------------------------------------------------------------------------------
LR test vs. logistic model: chibar2(01) = 23.47       Prob >= chibar2 = 0.0000
(est2 stored)

. eststo:  xtmelogit pursueonly W_SystemRELa persdumjlw_lag  land   swt _spline1 _spline2 _spline3  || year: W_SystemRELa, intpo
> ints(10)

Refining starting values: 

Iteration 0:   log likelihood = -230.79838  (not concave)
Iteration 1:   log likelihood = -227.48577  
Iteration 2:   log likelihood = -225.95219  

Performing gradient-based optimization: 

Iteration 0:   log likelihood = -225.95219  (not concave)
Iteration 1:   log likelihood = -225.93482  (backed up)
Iteration 2:   log likelihood = -225.88835  
Iteration 3:   log likelihood = -225.86705  
Iteration 4:   log likelihood = -225.84175  
Iteration 5:   log likelihood = -225.82273  
Iteration 6:   log likelihood = -225.81337  
Iteration 7:   log likelihood = -225.81125  
Iteration 8:   log likelihood = -225.81101  
Iteration 9:   log likelihood = -225.81099  

Mixed-effects logistic regression               Number of obs     =      5,338
Group variable: year                            Number of groups  =         55

                                                Obs per group:
                                                              min =         31
                                                              avg =       97.1
                                                              max =        141

Integration points =  10                        Wald chi2(7)      =     323.90
Log likelihood = -225.81099                     Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
    pursueonly |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
  W_SystemRELa |  -65.74841    17.5252    -3.75   0.000    -100.0972   -31.39964
persdumjlw_lag |   2.028426   .3655715     5.55   0.000     1.311919    2.744933
          land |   .2641475    .059654     4.43   0.000     .1472278    .3810672
           swt |  -2.547337   .2431913   -10.47   0.000    -3.023983    -2.07069
      _spline1 |  -.0208235   .0024085    -8.65   0.000    -.0255441   -.0161028
      _spline2 |   .0124326   .0016117     7.71   0.000     .0092737    .0155914
      _spline3 |  -.0029976   .0005644    -5.31   0.000    -.0041038   -.0018915
         _cons |   13.78476   3.675766     3.75   0.000     6.580395    20.98913
--------------------------------------------------------------------------------

------------------------------------------------------------------------------
  Random-effects Parameters  |   Estimate   Std. Err.     [95% Conf. Interval]
-----------------------------+------------------------------------------------
year: Independent            |
                sd(W_Syst~a) |   3.629941   .7816139      2.380211    5.535844
                   sd(_cons) |   .0001494   1.644888             0           .
------------------------------------------------------------------------------
LR test vs. logistic model: chi2(2) = 23.67               Prob > chi2 = 0.0000

Note: LR test is conservative and provided only for reference.
(est3 stored)

. eststo:  xtmelogit pursueonly reg5_w       persdumjlw_lag  land   swt _spline1 _spline2 _spline3  || EUGENERegions:, intpoints
> (10)

Refining starting values: 

Iteration 0:   log likelihood = -225.62833  (not concave)
Iteration 1:   log likelihood = -221.41703  (not concave)
Iteration 2:   log likelihood = -221.07381  

Performing gradient-based optimization: 

Iteration 0:   log likelihood = -221.07381  (not concave)
Iteration 1:   log likelihood = -219.96469  (not concave)
Iteration 2:   log likelihood = -219.13454  (not concave)
Iteration 3:   log likelihood = -218.60421  (not concave)
Iteration 4:   log likelihood = -218.19874  (not concave)
Iteration 5:   log likelihood = -217.98641  (not concave)
Iteration 6:   log likelihood = -217.69532  
Iteration 7:   log likelihood = -217.27139  (not concave)
Iteration 8:   log likelihood = -217.25637  
Iteration 9:   log likelihood = -216.98878  
Iteration 10:  log likelihood = -216.96342  
Iteration 11:  log likelihood = -216.96324  
Iteration 12:  log likelihood = -216.96324  

Mixed-effects logistic regression               Number of obs     =      5,324
Group variable: EUGENERegions                   Number of groups  =          5

                                                Obs per group:
                                                              min =        481
                                                              avg =    1,064.8
                                                              max =      1,454

Integration points =  10                        Wald chi2(7)      =     314.96
Log likelihood = -216.96324                     Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
    pursueonly |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
        reg5_w |  -13.15517   4.810805    -2.73   0.006    -22.58417   -3.726165
persdumjlw_lag |   1.515892   .3767437     4.02   0.000     .7774876    2.254296
          land |   .2989994   .0602412     4.96   0.000     .1809289    .4170699
           swt |  -2.195854   .2249645    -9.76   0.000    -2.636777   -1.754932
      _spline1 |  -.0186638   .0022768    -8.20   0.000    -.0231262   -.0142013
      _spline2 |   .0114186   .0015553     7.34   0.000     .0083702     .014467
      _spline3 |   -.002884   .0005647    -5.11   0.000    -.0039908   -.0017771
         _cons |   3.913604   2.279039     1.72   0.086      -.55323    8.380437
--------------------------------------------------------------------------------

------------------------------------------------------------------------------
  Random-effects Parameters  |   Estimate   Std. Err.     [95% Conf. Interval]
-----------------------------+------------------------------------------------
EUGENERegi~s: Identity       |
                   sd(_cons) |   2.470172   1.050246      1.073547    5.683732
------------------------------------------------------------------------------
LR test vs. logistic model: chibar2(01) = 70.46       Prob >= chibar2 = 0.0000
(est4 stored)

. eststo:  xtmelogit pursueonly reg5_w       persdumjlw_lag  land   swt _spline1 _spline2 _spline3  || EUGENERegions: reg5_w, in
> tpoints(10)

Refining starting values: 

Iteration 0:   log likelihood =   -225.262  (not concave)
Iteration 1:   log likelihood = -224.21399  (not concave)
Iteration 2:   log likelihood = -221.19399  

Performing gradient-based optimization: 

Iteration 0:   log likelihood = -221.19399  (not concave)
Iteration 1:   log likelihood = -215.93816  (not concave)
Iteration 2:   log likelihood = -214.86026  
Iteration 3:   log likelihood = -214.05406  
Iteration 4:   log likelihood = -213.47312  
Iteration 5:   log likelihood = -213.35372  
Iteration 6:   log likelihood =  -213.3495  
Iteration 7:   log likelihood = -213.34949  

Mixed-effects logistic regression               Number of obs     =      5,324
Group variable: EUGENERegions                   Number of groups  =          5

                                                Obs per group:
                                                              min =        481
                                                              avg =    1,064.8
                                                              max =      1,454

Integration points =  10                        Wald chi2(7)      =     318.90
Log likelihood = -213.34949                     Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
    pursueonly |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
        reg5_w |  -19.49707   5.850232    -3.33   0.001    -30.96331   -8.030823
persdumjlw_lag |   1.531073   .3820021     4.01   0.000     .7823628    2.279784
          land |   .2838577   .0604371     4.70   0.000     .1654031    .4023122
           swt |  -2.185153   .2246852    -9.73   0.000    -2.625528   -1.744778
      _spline1 |   -.018529    .002271    -8.16   0.000    -.0229802   -.0140778
      _spline2 |   .0113379   .0015492     7.32   0.000     .0083014    .0143743
      _spline3 |  -.0028678   .0005615    -5.11   0.000    -.0039683   -.0017674
         _cons |   6.135464   1.956974     3.14   0.002     2.299866    9.971062
--------------------------------------------------------------------------------

------------------------------------------------------------------------------
  Random-effects Parameters  |   Estimate   Std. Err.     [95% Conf. Interval]
-----------------------------+------------------------------------------------
EUGENERegi~s: Independent    |
                  sd(reg5_w) |    6.34979   2.598324      2.847404    14.16021
                   sd(_cons) |   1.10e-08   1.166444             0           .
------------------------------------------------------------------------------
LR test vs. logistic model: chi2(2) = 77.69               Prob > chi2 = 0.0000

Note: LR test is conservative and provided only for reference.
(est5 stored)

.  cd "C:\Users\Dell\Desktop"
C:\Users\Dell\Desktop

. esttab, b(a2) se(2) replace label star(* 0.10 ** 0.05 *** 0.01) scalars(N chi2  ll ) pr2(4) varwidth(25) modelwidth(9)  ///
> title(Table 2a: Uncertainty and Nuclear Proliferation, Multilevel Model)       ///
> nonumbers   ///
> mtitles("Model SW1" "Model SW2" "Model SW3" "Model SW4" "Model SW5")  ///
> order(W_SystemRELa reg5_w persdumjlw_lag) ///
> drop( _spline1 _spline2 _spline3) ///
> nogap

Table 2a: Uncertainty and Nuclear Proliferation, Multilevel Model
------------------------------------------------------------------------------------------
                          Model SW1    Model SW2    Model SW3    Model SW4    Model SW5   
------------------------------------------------------------------------------------------
eq1                                                                                       
W_SystemRELa                               -66.5***     -65.7***                          
                                         (17.19)      (17.53)                             
reg5_w                                                               -13.2***     -19.5***
                                                                    (4.81)       (5.85)   
persdumjlw_lag                 1.85***      2.02***      2.03***      1.52***      1.53***
                             (0.36)       (0.36)       (0.37)       (0.38)       (0.38)   
land                           0.25***      0.26***      0.26***      0.30***      0.28***
                             (0.06)       (0.06)       (0.06)       (0.06)       (0.06)   
Time since last pursueo~y     -2.63***     -2.55***     -2.55***     -2.20***     -2.19***
                             (0.25)       (0.24)       (0.24)       (0.22)       (0.22)   
Constant                      0.028         14.0***      13.8***      3.91*        6.14***
                             (0.33)       (3.61)       (3.68)       (2.28)       (1.96)   
------------------------------------------------------------------------------------------
lns1_1_1                                                                                  
Constant                    -0.0100        -0.27         1.29***      0.90**       1.85***
                             (0.19)       (0.22)       (0.22)       (0.43)       (0.41)   
------------------------------------------------------------------------------------------
lns1_1_2                                                                                  
Constant                                                -8.81                     -18.3   
                                                    (11012.62)                 (1.06e+08)   
------------------------------------------------------------------------------------------
Observations                   5338         5338         5338         5324         5324   
Pseudo R-squared                                                                          
chi2                          326.3        323.2        323.9        315.0        318.9   
ll                           -232.8       -225.9       -225.8       -217.0       -213.3   
------------------------------------------------------------------------------------------
Standard errors in parentheses
* p<0.10, ** p<0.05, *** p<0.01

. 
. rename _spline1 SW_spline1 

. rename _spline2 SW_spline2 

. rename _spline3 SW_spline3 

. 
. 
. *********************************************
. ******APPENDIX TABLE 11********************** - OK
. *********************************************
. *Uncertainty and Nuclear Proliferation, Gartzke and Jo | Hierarchical Model
. 
. btscs gjprog  year ccode, g(gjt) nspline(3)

. eststo clear

. eststo:  xtmelogit gjprog              persdumjlw_lag  land   gjt _spline1 _spline2 _spline3  || year:, intpoints(10)

Refining starting values: 

Iteration 0:   log likelihood = -255.79636  (not concave)
Iteration 1:   log likelihood = -252.38951  (not concave)
Iteration 2:   log likelihood = -252.23808  

Performing gradient-based optimization: 

Iteration 0:   log likelihood = -252.23808  (not concave)
Iteration 1:   log likelihood = -251.73642  
Iteration 2:   log likelihood = -251.06943  
Iteration 3:   log likelihood = -251.05571  
Iteration 4:   log likelihood = -251.05566  
Iteration 5:   log likelihood = -251.05566  

Mixed-effects logistic regression               Number of obs     =      5,337
Group variable: year                            Number of groups  =         55

                                                Obs per group:
                                                              min =         31
                                                              avg =       97.0
                                                              max =        141

Integration points =  10                        Wald chi2(6)      =     325.08
Log likelihood = -251.05566                     Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
        gjprog |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
persdumjlw_lag |   2.153785   .3763721     5.72   0.000     1.416109    2.891461
          land |    .299838   .0588042     5.10   0.000     .1845839    .4150921
           gjt |  -2.178255   .1932108   -11.27   0.000    -2.556941   -1.799569
      _spline1 |  -.0171946   .0020184    -8.52   0.000    -.0211506   -.0132385
      _spline2 |    .010045   .0014873     6.75   0.000       .00713      .01296
      _spline3 |  -.0022442    .000647    -3.47   0.001    -.0035123   -.0009761
         _cons |  -.3110462   .3189233    -0.98   0.329    -.9361243    .3140319
--------------------------------------------------------------------------------

------------------------------------------------------------------------------
  Random-effects Parameters  |   Estimate   Std. Err.     [95% Conf. Interval]
-----------------------------+------------------------------------------------
year: Identity               |
                   sd(_cons) |   .8707106   .1701056      .5937171    1.276933
------------------------------------------------------------------------------
LR test vs. logistic model: chibar2(01) = 28.42       Prob >= chibar2 = 0.0000
(est1 stored)

. eststo:  xtmelogit gjprog W_SystemRELa persdumjlw_lag  land   gjt _spline1 _spline2 _spline3  || year:, intpoints(10)

Refining starting values: 

Iteration 0:   log likelihood = -248.33899  (not concave)
Iteration 1:   log likelihood = -244.40953  
Iteration 2:   log likelihood = -243.33713  

Performing gradient-based optimization: 

Iteration 0:   log likelihood = -243.33713  
Iteration 1:   log likelihood = -243.30771  
Iteration 2:   log likelihood = -243.30753  
Iteration 3:   log likelihood = -243.30753  

Mixed-effects logistic regression               Number of obs     =      5,335
Group variable: year                            Number of groups  =         55

                                                Obs per group:
                                                              min =         31
                                                              avg =       97.0
                                                              max =        141

Integration points =  10                        Wald chi2(7)      =     324.70
Log likelihood = -243.30753                     Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
        gjprog |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
  W_SystemRELa |  -61.05783   14.94707    -4.08   0.000    -90.35354   -31.76211
persdumjlw_lag |   2.345031   .3887324     6.03   0.000      1.58313    3.106933
          land |   .3223537   .0585373     5.51   0.000     .2076227    .4370846
           gjt |  -2.116698   .1866109   -11.34   0.000    -2.482448   -1.750947
      _spline1 |  -.0162603   .0019473    -8.35   0.000     -.020077   -.0124437
      _spline2 |   .0093182   .0014514     6.42   0.000     .0064736    .0121629
      _spline3 |   -.001977   .0006492    -3.05   0.002    -.0032495   -.0007045
         _cons |   12.43569   3.150641     3.95   0.000     6.260542    18.61083
--------------------------------------------------------------------------------

------------------------------------------------------------------------------
  Random-effects Parameters  |   Estimate   Std. Err.     [95% Conf. Interval]
-----------------------------+------------------------------------------------
year: Identity               |
                   sd(_cons) |   .6550621   .1574262      .4089965    1.049169
------------------------------------------------------------------------------
LR test vs. logistic model: chibar2(01) = 13.98       Prob >= chibar2 = 0.0001
(est2 stored)

. eststo:  xtmelogit gjprog W_SystemRELa persdumjlw_lag  land   gjt _spline1 _spline2 _spline3  || year: W_SystemRELa, intpoints
> (10)

Refining starting values: 

Iteration 0:   log likelihood = -248.64675  (not concave)
Iteration 1:   log likelihood =  -244.2289  (not concave)
Iteration 2:   log likelihood = -244.06639  

Performing gradient-based optimization: 

Iteration 0:   log likelihood = -244.06639  
Iteration 1:   log likelihood = -243.28288  (not concave)
Iteration 2:   log likelihood = -243.27945  (backed up)
Iteration 3:   log likelihood = -243.27095  
Iteration 4:   log likelihood =  -243.2464  
Iteration 5:   log likelihood =  -243.2409  
Iteration 6:   log likelihood = -243.23724  
Iteration 7:   log likelihood = -243.23664  
Iteration 8:   log likelihood = -243.23658  
Iteration 9:   log likelihood = -243.23658  

Mixed-effects logistic regression               Number of obs     =      5,335
Group variable: year                            Number of groups  =         55

                                                Obs per group:
                                                              min =         31
                                                              avg =       97.0
                                                              max =        141

Integration points =  10                        Wald chi2(7)      =     325.38
Log likelihood = -243.23658                     Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
        gjprog |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
  W_SystemRELa |  -60.19371   15.21563    -3.96   0.000    -90.01579   -30.37162
persdumjlw_lag |   2.353837    .389624     6.04   0.000     1.590188    3.117486
          land |   .3231359    .058607     5.51   0.000     .2082683    .4380035
           gjt |  -2.114915   .1859916   -11.37   0.000    -2.479452   -1.750378
      _spline1 |  -.0162438   .0019408    -8.37   0.000    -.0200477     -.01244
      _spline2 |   .0093083    .001447     6.43   0.000     .0064722    .0121444
      _spline3 |  -.0019746   .0006479    -3.05   0.002    -.0032445   -.0007047
         _cons |   12.24834   3.199746     3.83   0.000     5.976957    18.51973
--------------------------------------------------------------------------------

------------------------------------------------------------------------------
  Random-effects Parameters  |   Estimate   Std. Err.     [95% Conf. Interval]
-----------------------------+------------------------------------------------
year: Independent            |
                sd(W_Syst~a) |   3.097779    .743255      1.935621    4.957704
                   sd(_cons) |   .0000514   1.722417             0           .
------------------------------------------------------------------------------
LR test vs. logistic model: chi2(2) = 14.12               Prob > chi2 = 0.0009

Note: LR test is conservative and provided only for reference.
(est3 stored)

. eststo:  xtmelogit gjprog reg5_w       persdumjlw_lag  land   gjt _spline1 _spline2 _spline3  || EUGENERegions:, intpoints(10)

Refining starting values: 

Iteration 0:   log likelihood = -254.90001  (not concave)
Iteration 1:   log likelihood = -251.27839  
Iteration 2:   log likelihood = -242.76579  

Performing gradient-based optimization: 

Iteration 0:   log likelihood = -242.76579  
Iteration 1:   log likelihood = -237.18207  (not concave)
Iteration 2:   log likelihood = -237.05604  
Iteration 3:   log likelihood = -236.65969  
Iteration 4:   log likelihood = -236.53566  
Iteration 5:   log likelihood = -236.53386  
Iteration 6:   log likelihood = -236.53386  

Mixed-effects logistic regression               Number of obs     =      5,321
Group variable: EUGENERegions                   Number of groups  =          5

                                                Obs per group:
                                                              min =        478
                                                              avg =    1,064.2
                                                              max =      1,454

Integration points =  10                        Wald chi2(7)      =     328.87
Log likelihood = -236.53386                     Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
        gjprog |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
        reg5_w |  -25.33985    4.64888    -5.45   0.000    -34.45149   -16.22822
persdumjlw_lag |   2.228181     .39459     5.65   0.000     1.454799    3.001563
          land |    .286198   .0567157     5.05   0.000     .1750373    .3973587
           gjt |  -1.859241   .1778157   -10.46   0.000    -2.207753   -1.510729
      _spline1 |  -.0145482   .0018846    -7.72   0.000    -.0182418   -.0108545
      _spline2 |   .0084113   .0014189     5.93   0.000     .0056303    .0111923
      _spline3 |  -.0017872   .0006412    -2.79   0.005    -.0030438   -.0005305
         _cons |   8.748124   2.472188     3.54   0.000     3.902725    13.59352
--------------------------------------------------------------------------------

------------------------------------------------------------------------------
  Random-effects Parameters  |   Estimate   Std. Err.     [95% Conf. Interval]
-----------------------------+------------------------------------------------
EUGENERegi~s: Identity       |
                   sd(_cons) |   3.489579   1.398202      1.591162    7.653001
------------------------------------------------------------------------------
LR test vs. logistic model: chibar2(01) = 56.20       Prob >= chibar2 = 0.0000
(est4 stored)

. eststo:  xtmelogit gjprog reg5_w       persdumjlw_lag  land   gjt _spline1 _spline2 _spline3  || EUGENERegions: reg5_w, intpoi
> nts(10)

Refining starting values: 

Iteration 0:   log likelihood = -254.26829  (not concave)
Iteration 1:   log likelihood = -252.06435  (not concave)
Iteration 2:   log likelihood =  -243.5959  

Performing gradient-based optimization: 

Iteration 0:   log likelihood =  -243.5959  (not concave)
Iteration 1:   log likelihood = -234.65356  (not concave)
Iteration 2:   log likelihood = -232.82141  (not concave)
Iteration 3:   log likelihood = -232.57932  
Iteration 4:   log likelihood = -231.67306  
Iteration 5:   log likelihood = -231.65825  
Iteration 6:   log likelihood = -231.65824  

Mixed-effects logistic regression               Number of obs     =      5,321
Group variable: EUGENERegions                   Number of groups  =          5

                                                Obs per group:
                                                              min =        478
                                                              avg =    1,064.2
                                                              max =      1,454

Integration points =  10                        Wald chi2(7)      =     332.66
Log likelihood = -231.65824                     Prob > chi2       =     0.0000

--------------------------------------------------------------------------------
        gjprog |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
        reg5_w |  -30.12797   6.164905    -4.89   0.000    -42.21097   -18.04498
persdumjlw_lag |   2.235329   .4028527     5.55   0.000     1.445752    3.024906
          land |   .2690706   .0571884     4.70   0.000     .1569833    .3811578
           gjt |   -1.83036   .1766071   -10.36   0.000    -2.176504   -1.484216
      _spline1 |  -.0142368   .0018766    -7.59   0.000    -.0179148   -.0105588
      _spline2 |   .0082133   .0014173     5.80   0.000     .0054354    .0109912
      _spline3 |  -.0017339   .0006445    -2.69   0.007    -.0029972   -.0004706
         _cons |   10.19026   1.860336     5.48   0.000     6.544067    13.83645
--------------------------------------------------------------------------------

------------------------------------------------------------------------------
  Random-effects Parameters  |   Estimate   Std. Err.     [95% Conf. Interval]
-----------------------------+------------------------------------------------
EUGENERegi~s: Independent    |
                  sd(reg5_w) |   8.085112   3.196652      3.725136    17.54809
                   sd(_cons) |   4.34e-08   1.317994             0           .
------------------------------------------------------------------------------
LR test vs. logistic model: chi2(2) = 65.95               Prob > chi2 = 0.0000

Note: LR test is conservative and provided only for reference.
(est5 stored)

. esttab, b(a2) se(2) replace label star(* 0.10 ** 0.05 *** 0.01) scalars(N chi2  ll ) pr2(4) varwidth(25) modelwidth(9)  ///
> title(Table 2b: Uncertainty and Nuclear Proliferation, Multilevel Model)       ///
> nonumbers   ///
> mtitles("Model SW1" "Model SW2" "Model SW3" "Model SW4" "Model SW5")  ///
> order(W_SystemRELa reg5_w persdumjlw_lag) ///
> drop( _spline1 _spline2 _spline3) ///
> nogap

Table 2b: Uncertainty and Nuclear Proliferation, Multilevel Model
------------------------------------------------------------------------------------------
                          Model SW1    Model SW2    Model SW3    Model SW4    Model SW5   
------------------------------------------------------------------------------------------
eq1                                                                                       
W_SystemRELa                               -61.1***     -60.2***                          
                                         (14.95)      (15.22)                             
reg5_w                                                               -25.3***     -30.1***
                                                                    (4.65)       (6.16)   
persdumjlw_lag                 2.15***      2.35***      2.35***      2.23***      2.24***
                             (0.38)       (0.39)       (0.39)       (0.39)       (0.40)   
land                           0.30***      0.32***      0.32***      0.29***      0.27***
                             (0.06)       (0.06)       (0.06)       (0.06)       (0.06)   
Time since last gjprog        -2.18***     -2.12***     -2.11***     -1.86***     -1.83***
                             (0.19)       (0.19)       (0.19)       (0.18)       (0.18)   
Constant                      -0.31         12.4***      12.2***      8.75***      10.2***
                             (0.32)       (3.15)       (3.20)       (2.47)       (1.86)   
------------------------------------------------------------------------------------------
lns1_1_1                                                                                  
Constant                      -0.14        -0.42*        1.13***      1.25***      2.09***
                             (0.20)       (0.24)       (0.24)       (0.40)       (0.40)   
------------------------------------------------------------------------------------------
lns1_1_2                                                                                  
Constant                                                -9.88                     -17.0   
                                                    (33522.68)                 (30377664.60)   
------------------------------------------------------------------------------------------
Observations                   5337         5335         5335         5321         5321   
Pseudo R-squared                                                                          
chi2                          325.1        324.7        325.4        328.9        332.7   
ll                           -251.1       -243.3       -243.2       -236.5       -231.7   
------------------------------------------------------------------------------------------
Standard errors in parentheses
* p<0.10, ** p<0.05, *** p<0.01

. log close
      name:  <unnamed>
       log:  C:\Users\Dell\Dropbox\OutcomeUncertainty\II Replication Folder\Nuclear Proliferation Log.log
  log type:  text
 closed on:  24 Sep 2017, 01:00:38
--------------------------------------------------------------------------------------------------------------------------------
